PWM Noise/whine while on motor; why human ear hear them

Thread Starter

Vindhyachal Takniki

Joined Nov 3, 2014
594
When I drive motor with PWM from MCU I heard a whine sound. Several people suggest to increase the PWM freq above audible range of human i.e 20Khz.
But one thing I dont understand human ear heard sound when there is compression/rarefaction around local region.
But PWM freq let say at 5Khz is electrical freq. It is not making any compression/rarefaction nearby. Then why this noise & how do human ear heard these?

Then if I rum MCU core at 10Khz low freq then it should also make noise? But when I do it, it dont make any noise
 

crutschow

Joined Mar 14, 2008
34,285
The noise you hear is not directly from the electrical signal, it's from the mechanical parts (particularly the magnetic laminations) moving slightly in response to the magnetic field from the electrical current moving through the magnetic material.
The MCU has no magnetics so there is nothing move due to the electrical signal.
 

dannyf

Joined Sep 13, 2015
2,197
"PWM Noise/whine while on motor; why human ear hear them"

I guess hearing noises or sound is exactly what your ears are designed to do?
 

GopherT

Joined Nov 23, 2012
8,009
"PWM Noise/whine while on motor; why human ear hear them"

I guess hearing noises or sound is exactly what your ears are designed to do?
I guess big text makes you feel like you are answering the question - you'll have to read more than an attention-getting headline to understand the OP's question.
 

Tonyr1084

Joined Sep 24, 2015
7,853
The reason why you hear the noise at 5KhZ and not at 10KhZ is because of harmonics. All things vibrate at a natural rate - whatever that rate may be. Have you ever seen someone shatter a glass using just their voice? It's because they hit the pitch that makes the glass ring at its natural resonance frequency. Slightly above or below the resonance and it won't vibrate nearly as much or as hard. So at 5KhZ you must be close to the motors natural resonance, causing it to make a physical sound. 10KhZ is just not the right resonant frequency for the motor to make noise (that you can hear).
 

GopherT

Joined Nov 23, 2012
8,009
Have you ever seen someone shatter a glass using just their voice? It's because they hit the pitch that makes the glass ring at its natural resonance frequency.
Have you seen this done? No, it cannot be done with an unamplified voice. Also, it is not likely to be done with an amplified voice - most likely demonstrated with a tone generator. Very few humans can sustain a perfect resonance note long enough (and keep the glass in resonance long enough) to cause material fatigue and shatter the glass - that is, I think you are "mis-remembering" the old TV commercial - the speaker was next to the glass and they ask, "is it live or is it Memorex" - to mean the Tape can retain he high notes of a live performance.

Slightly above or below the resonance and it won't vibrate nearly as much or as hard. So at 5KhZ you must be close to the motors natural resonance, causing it to make a physical sound. 10KhZ is just not the right resonant frequency for the motor to make noise (that you can hear).
@Vindhyachal Takniki
The key to the hearing 5kHz and not hearing 10kHz in this application is that human hearing decreases in sensitivity as frequency increases AND the windings in the motor are an inductor and attenuate the signal more and more as frequency increases.
 

Tonyr1084

Joined Sep 24, 2015
7,853
Have you seen this done?
Not in person, but I have seen it on TV at least one man who could shatter a wine glass using just his vocal cords (think it was Mythbusters). Placing a soda straw in the glass lets him know when he has the right pitch. And yes, I've seen it done.

As for my comment on resonance, I've done something similar (to the glass): In a bathroom where the walls are square and the person sitting - um, just leave it at sitting for now - can hum a certain frequency in which the sound will reverberate off the walls at its NATURAL frequency. The resonance will seemingly cancel out most other noises. (Personal experience)

You have a different opinion (or consideration or knowledge) on why a motor will make more noise at 5KhZ as opposed to 10KhZ. I accept your thought on that. However, I still remain of the same opinion that a motor (such as a small motor) will make more noise at a higher frequency than a larger motor will. The amount of mass that must go into resonance is greater, therefore has a longer period, thus, a larger motor will whine at a lower frequency.

I have several treadmill motors that operate on PWM supplied power. At their lower RPM's they make a lot more noise because of the slow rate of switching (actually the low percentage of duty frequency) - and their greater mass than - oh, lets say - a hair dryer motor, which if powered from PWM would resonate at a much higher frequency. This is what I believe. If you can show me something different (aside from an opinion) then I may change MY beliefs. Hey! I'm open to learning new things. Just takes reason. If it's sound then I'll accept it.

Oh, and the glass on Mythbusters: Found it.
 
Last edited:

IamJatinah

Joined Oct 22, 2014
136
Hi All,

If we recall Mr Tesla and his "vibration machine" which also worked on resonance of structures, there are a few things I have seen in the thread to touch on....

The thought of the vibrating "wires" on wrapped transformers can indeed be a factor, as is something "Inside" the ferrite materials. Science has found some things, and yep, even MythBusters has proven vibration-resonances and the destructive forces which can be obtained within a medium, say glass.

Ferrite materials most often have molecules in disarray when not energized, then when energized, these molecules "snap" into an alignment, very uniformly and abruptly, similar to quenching steels. This also adds to audible noise of transformers, along with the winding factors or movement. Also, "Q" of a transformer or it's ferro-resonant center will produce the best coupling with the least noise, SWR, reactance issues, if driven at it's favorite frequency.
For Example, say I have a switching circuit operating at 25khz, and a transformer designed to be switched at 39khz I now have some issues to contend with as a designer/servicer as the drive frequency doesn't match the "Q" and may aid or oppose harmonics produced and cause SWR issues.

All Ferrite materials will allow a buildup of external FLUX which is magnetic in nature and does impart forces onto other nearby parts, bits and pieces.

There are also inductive forces at many components in heavy current rails, and when one hits a full load situation with any electrical device, one can hear things getting loose, or near possible trouble.

Let's also remember simply running a current thru a piece of wire will cause a slight motive movement of that wire due to initial generated flux and external forces close to that wire.

Normally, the tighter the designs, the quieter the machine, but there can be many factors....

;o)
 
Top